1. Field of the Invention
This invention relates to real-time messaging technologies, user interfaces for such technologies, and the ability to recall and share previously conducted conversations.
2. Background of the Invention
Real-time collaboration tools such as IBM's Same Time, MSN's Messenger, Yahoo! Messenger, and AOL Instant Messenger (“IM”)™ segment conversations by assigning break points to signify the end of conversations. Methods used to determine a break point in the thread of conversation include one or more of the following:                (a) user interface or chat window closure;        (b) logoff of the user from the messaging service;        (c) exiting the messaging client program; and        (d) shutting down the client system's operating system or turning off the client device.        
These break points affect both the logged transcript of the conversation held between the conversants and, perhaps more importantly, what happens when communications between the same conversants is resumed after the break point is set.
For example, after a messaging session between a first, a second and a third user, a break point is set when the first user closes his or her client messaging window. Upon resuming conversation by the first user with the second or third user, or with both, the messaging log shows a new conversation and, more importantly, the previous entries from the previously concluded conversation are not displayed.
If the first user wishes to refresh his or her memory as to what was said during the previously concluded conversation, or where the thoughts ended during the conversation, he or she must find and review transcripts of many previously completed “chats”, or more often, he or she will ask the other conversants questions about what was said during the previous conversation. This may require several exchanges between the conversants to establish a “context” of the conversation, and, depending on the accuracy of the memories of the other conversants, may lead to repeating subjects or inaccurately recalling previous discussions. Often, this is very frustrating to the participants as well.
This limitation of the known real-time messaging technology limits the effectiveness of the tools. A fundamental flaw with automatically assigning break points in real-time conversation transcripts is that a conversation may not be over just because the user closes a window, shuts down their computer, etc. These events are only logically associated with the end of a logical conversation in some situations, but not always and not reliably.
Now consider another example which illustrates the same problem:                (1) a first user closes his or her messaging client user interface window, which terminates the transcript on the first user's client device;        (2) perhaps 45 minutes later, the first user restarts his or her messaging client and is ready to receive messages, albeit the messaging client has started a new conversation transcript;        (3) another participant in the original conversation of (1) sends an entry related to the original conversation, but after the first user has restarted his or her messaging client; and        (4) the new entry is received by the first user but displayed in the context of the new conversation, out of context with the original conversation.        
In this scenario, which is not uncommon, the first user has mentally lost the context of the original conversation, especially if other conversations with other conversants or about other subjects were conducted during the meantime.
Consider another example, similar to the previous example, but in a situation where one conversant sends a message to another, but then has to shut down or reboot his or her messaging device. The recipient of the question replies “Sure” or “Sounds OK to me”, but this message is not received until the originator of the question reboots the messaging client. Unfortunately, when the reply is received by the first user, the message arrives out of context, and it is unclear to the first user what the question was.
Some conversations, even though the short term exchanges are made in real-time, actually evolve and develop over hours or even days, as some of the participants are drawn away for other matters, or some of the questions posed require time to consider or research, etc. This situation, which is also very common, leaves all users with fragments of conversation threads, none of which can be easily reconstructed into a single context of conversation.
Current solutions to help deal with this problem are manual. A user may try to keep chat windows open as long as possible, they may look up and visually search through many previous chat transcripts, they may ask other participants for context reminders, or they may rely on their best memory of what was discussed. In all of these solutions, the accuracy of the context recall is less than desired, the process may be frustrating, and much of the benefit of real-time communications is lost on inefficiency of context recollection.